2
$\begingroup$

Currently, I am facing this problem:

Given two real functions $A( \vec x )$ and $B( \vec x ):\Bbb R^N\to \Bbb R$, I want to find a third real, monotonic function $f(x):\Bbb R\to\Bbb R$ such that:

$$A\big(f( \vec x )\big)=f\big(B( \vec x )\big)$$

where the simplified the notation writing $f( \vec x )$ means $$ f(x_1,x_2,\ldots,x_n) = \big(f(x_1), f(x_2),\ldots, f(x_n)\big). $$

I am interested in either having a formula/method for finding $f$, or even just having a proof that $f$ exists (or doesn't) under some specific conditions. Eventually, I am interested also in the solution in the case $N=1$.

Also, does this type of problem have a specific name?

Thank you very much!

$\endgroup$
  • $\begingroup$ Fixed! I hope now it's fun enough $\endgroup$ – SchroedingerDidStuff Feb 10 at 9:37
  • 2
    $\begingroup$ you might also want to consider changing the name of your profile; the adjectives you use to characterise Schrödinger are inappropriate, IMO. $\endgroup$ – Carlo Beenakker Feb 10 at 9:49
  • 1
    $\begingroup$ Ok, I also changed the nick to a more appropriate one. I guess the next answer I'll receive is not gonna be on the question I asked, but on the way, I use the comma. Maybe I should ask this same question on a grammar forum... $\endgroup$ – SchroedingerDidStuff Feb 10 at 9:58
  • 1
    $\begingroup$ It seems too easy: if $A(x_1,x_2)=1+x_1$ and $B(x_1,x_2)=x_1+x_2$ then you are looking for $f$ such that $1+f(x_1)=f(x_1+x_2)$ which cannot be, for example if $x_2=0$. Maybe you missed some conditions? (Or did I?) $\endgroup$ – Yaakov Baruch Feb 10 at 10:09
  • 3
    $\begingroup$ Also, does this type of problem have a specific name? Given two maps $A$ and $B$, trying to find an $f$ such that $A \circ f =f \circ B$ is a problem of semi-conjugacy. $\endgroup$ – Laurent Berger Feb 10 at 13:12
5
$\begingroup$

Existence of such $f$ is a very strong condition on $A$ and $B$ which is called semi-conjugacy. Of course, for generic $A$ and $B$ function $f$ does not exist. Suppose for simplicity that $N=1$. Then it is clear that the image of any fixed point of $B$ under $f$ is a fixed point of $A$. Furthermore, when $N=1$, your equation implies that $$A^n\circ f=f\circ B^n$$ where $A^n$ means the $n$-th iterate. This implies that periodic points of $B$ are mapped to periodic points of $A$ of the same period. So we have some very complicated relation between $A$ and $B$.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Alexandre, thank you very much! I think I need to study a little more the theory of topological conjugacy, as there are some points that I don't fully understand. By adding slightly more stringent conditions (i.e. A() and B() continuous and f derivable) is it possible to find a solution to the problem? Otherwise, what are the typical hypotheses to solve it? (if they exist) Eventually, could you suggest me a book or article that could help me in understanding better how to solve the problem? Again, thanks a lot! $\endgroup$ – SchroedingerDidStuff Feb 10 at 14:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.